396 A. Ljungqvist
Spindt (1989)framework, since public information is freely available and so there is no
need to compensate investors for it by leaving money on the table. Loughran and Ritter
prefer a behavioral explanation, which will be discussed more fully in Section6.3.In
short, when the IPO is doing poorly (and so the price is likely to be revised downwards),
issuers bargain hard with the underwriter over the issue price. When the IPO is doing
well (and so the price is likely to be revised upwards), issuers are complacent. This
leads to an asymmetric relation between prior market returns and offer price revisions,
at least to the extent that the state of the market correlates with how the IPO is doing.
Lowry and Schwert (2004)reexamine this question. While their findings confirm the
existence of a positive and statistically significant relation between offer price revisions
and pre-pricing market returns, they argue that this effect is negligible economically.
Edelen and Kadlec (2005), too, reexamineLoughran and Ritter’s (2002)critique, and
show that the apparent asymmetry may be driven by sample selection bias. In a sam-
ple ofcompletedIPOs, negative market returns have indeed no effect on offer price
revisions. But negative market returns have a significant impact on the decision to with-
draw the IPO. When this is taken into account using theHeckman (1979)approach, the
asymmetry disappears.
Whether symmetric or asymmetric, public information appears not to be fully priced.
Why not? In contrast toLoughran and Ritter (2002), Edelen and Kadlec (2005)propose
arationalexplanation, noting that issuers must trade off the proceeds from the IPO
against the probability of the IPO succeeding. In the context of a search model, aggres-
sive pricing increases the probability of failure. When comparable firms’ valuations are
low, the IPO is likely to generate relatively little ‘surplus’ for the issuer. Therefore, the
issuer has little to lose if the deal fails, and pushes the underwriter to extract as high
proceeds as possible, even though this implies a greater risk of the deal failing. When
comparable firm valuations are high, the issuer is unwilling to risk failure because there
is much to be gained from going public. In this situation, the issuer does not insist on ag-
gressive pricing. Thus as comparable firms’ valuations increase, so too does the degree
of underpricing.
3.3. Principal-agent models
Theories of bookbuilding stress the important role of investment banks in eliciting in-
formation that is valuable in price-setting, and the benefit of giving them discretion
over allocation decisions. Some authors—most prominently perhapsLoughran and Rit-
ter (2004)—stress the ‘dark side’ of these institutional arrangements, by highlighting
the potential for agency problems between the investment bank and the issuing firm.
A multitude of regulatory investigations following the bursting of the late 1990s
‘dot-com bubble’ has recently revived academic interest in agency models of IPO under-
pricing. For instance, the fact that underpricing represents a wealth transfer from the IPO
company to investors can give rise to rent-seeking behavior, whereby investors compete
for allocations of underpriced stock by offering the underwriter side-payments. Such
side-payments could take the form of excessive trading commissions paid on unrelated